During my time at the NIH National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK), I worked on a cohort study that followed 5,000 patients for 10 years. The findings challenged everything I learned in medical school about weight, BMI, and metabolic health.
I can't share the raw data — it's still under publication embargo — but I can share the conclusions. And they changed my practice.
The Study Design
5,000 adults, ages 30-65, recruited from primary care clinics across the US. Measured at baseline: BMI, waist circumference, body fat percentage (Bod Pod), fasting glucose, insulin, lipids, blood pressure, inflammatory markers, and a comprehensive metabolic panel.
Then followed for 10 years. Annual measurements. No intervention — this was observational. The goal was to see which baseline measures predicted metabolic outcomes: type 2 diabetes, cardiovascular events, fatty liver disease, and all-cause mortality.
The Surprising Findings
Finding 1: BMI was a weak predictor.
Baseline BMI predicted diabetes with an AUC of 0.62. That's barely better than a coin flip. Waist circumference predicted diabetes with an AUC of 0.71. Body fat percentage: 0.74. But the combination of waist + fasting insulin: 0.84.
The message was clear: BMI alone is insufficient. But we already knew that.
Finding 2: The "metabolically healthy obese" were real — but fragile.
About 18% of obese patients (BMI >30) were metabolically healthy at baseline: normal glucose, lipids, blood pressure, and inflammatory markers. But 10 years later, only 7% remained metabolically healthy. The rest had developed at least one metabolic abnormality.
So "metabolically healthy obese" exists. But it's not a stable state. Time and aging eventually catch up.
Finding 3: Weight fluctuation was worse than stable weight.
Patients who cycled weight — lost 20 pounds, gained 25, lost 15, gained 20 — had worse metabolic outcomes than patients who stayed consistently overweight. The yo-yo effect wasn't just psychologically damaging. It was metabolically damaging.
This was the finding that changed my practice. I stopped celebrating rapid weight loss. I started emphasizing sustainable, gradual changes.
What I Do Differently Now
I no longer set weight loss goals. I set behavior goals:
- Walk 30 minutes daily
- Resistance train twice weekly
- Eat protein at every meal
- Sleep 7-8 hours
- Manage stress (meditation, therapy, whatever works)
If weight loss happens as a side effect, great. If not, the metabolic improvements still occur. I've seen patients improve their glucose, lipids, and blood pressure with zero weight change. The scale is not the scoreboard.
The trend tracker on this site? I built it because I wanted patients to see that health metrics improve even when weight plateaus. The data supports it. My NIH data supports it. And after 15 years in practice, I've seen it hundreds of times.
— Dr. Chen
David Chen is a board-certified Internal Medicine physician. This article reflects clinical observations and personal experience. For medical advice, consult your healthcare provider.